Carbohydrates decorate proteins and cell surfaces serving as recognition entities for disease and immune processes, are implicated in cell aging and are used as biological signal carriers. One reason they are useful in so many ways is that they can adopt a staggering variety of shapes. Partly because of their importance to biological systems, and partly because of their structural complexity, it is critical that scientists, especially those studying their biological roles, be able easily to generate accurate and useful 3D models of them and their complexes with proteins. However, because of the structural complexity of carbohydrates, generation of these models is very difficult. Even researchers who are comfortable with computer modeling of biological systems can find the task daunting. It is not sufficient, however, only to facilitate accurate modeling for computational specialists. In order to overcome the barrier to adoption of such methods by a broad base of researchers, it is important that the software be intuitive and provides guidance in appropriate use. The GLYCAM-Web suite of online tools (www.glycam.org) has been making steps toward this goal for a number of years. Six years ago, it began as an in-house tool for the Woods carbohydrate modeling research group. Shortly thereafter it was released publically as an online utility free of charge. Site statistics have been gathered for approximately 5 years, and typically averages 400-500 visits each month. Based on support requests, solicited user letters, and references to the site found in academic journals, it is primarily used by persons engaged in biological research. The site was not initially designed with the anticipation of such expansion, nor was it initially intended for such a wide base of researchers. The purpose of this project is to update and transform the GLYCAM-Web suite of modeling tools into a stable, robust, informative and intuitive utility that is capable of meeting the current and future needs of glycobiology research. To achieve these goals, the site's software will be restructured so that it is built of self-contained, well-tested modules that are easy to maintain and modify. Its data structure will facilitate the addition of new functionality. To enable external researchers to use the software locally and modify it to meet their needs, the command-line interface on which the site is built will be released to the public. Submissions and corrections from the user community will be encouraged. To make the site usable by a wider variety of researchers, notably that are not primarily computer scientists, several features will be added or enhanced and site documentation will be greatly expanded based on user requests. Among the features to be added or enhanced are smarter error- checking, simplified builds of complex structures, an increased number of supported file formats, and interfaces with external web sites.